This research has two major purposes: The first purpose is to validate a technique to quantify personality characteristics which discriminate between opiate addicts and non-addicts -- specifically, the characteristics of "dependence" and "coolness" as they contribute to a propensity in an individual to take greater risks in a group than when acting alone. Validated, the technique may be used both as a screening device for treatment and a post-rehabilitation and follow-up measure. The second purpose is to substantiate the utility of, and to refine the design for, testing the effectiveness of a rehabilitation program which constitutes a modification of present programs -- specifically, intensive therapy for addicts in non-addict groups, after a preliminary period in programs of the current type. For the present study, five different types of groups (varying in ratio of addict to non-addict membership) will be used -- six groups of each type; six members (all male) per facility. Techniques devised for research on the "group induced shift" (i.e., that groups choose riskier alternatives than individuals) will be utilized. Each subject will fill out a risk-taking questionnaire three times: by himself (pre), in a groups discussion leading to consensus, then again by himself (post). The group consensus score will be compared to the averages (pre and post) of the individual scores, and the individual pre scores will be compared to the individual post scores. Tests for locus of control and ability to attend will also be administered, and group discussions will be taped for interaction process analysis. It is hypothesized that 1) addicts will choose a greater risk level on the pre-test than non-addicts; 2) addicts will show a greater shift toward risk than non-addicts; 3) shift toward risk will vary with the ratio of addict to non-addict group membership; and 4) addicts will show a lesser return on their post-test to pre-test scores.